MISCELLANEA. 
I. Fecundity of Swine*. 
By frank M. surface, Ph.D. 
During the course of certain investigations in the laboratory of the Maine Agricultural 
Experimental Station it became desirable for purposes of comparison to have data on other 
fecundity distributions than those already accessible in the literature. In this connection it 
was decided to fit frequency curves to the data published some time ago by Mr George Rommel t 
on the fecundity of brood sows. A large amount of data relating to the size of litters produced 
by registered sows of two breeds, Duroc Jersey and Poland China, had been extracted from the 
stud books and published by this writer. The raw material is given in the form of frequency 
distributions for single years. In the case of the Poland China sows the statistics were taken 
from two sources : viz., the American Poland China Record and the Ohio Poland China Record. 
In each case two five year periods, 1882-1886 and 1898-1902, were selected for comparison. 
Tables I, II, IV and V of the above-mentioned circular give the raw data for each year as 
obtained from the two sources. 
For the Duroc Jersey data the National Duroc Jersey Record was used. Tables VIII, IX and 
X give the number and size of litters registered for each of the fifteen consecutive years from 
1888 to 1902. In this case the earlier years contained relatively few records. 
In all cases however, and especially for the later years, the number of litters dealt with was 
very large. For this reason the statistics are of considerable value. No reductions other than 
the means for each year were given in the original paper. Later Rommel and Phillips J in a 
similar set of data for the year 1902 studied the inheritance of size of litter from Poland China 
sows. In the course of this work the standard deviations and coefficients of correlation were 
determined. The data with which the present paper deals have never been reduced although 
the variation constants in the case of Poland China sows are, as would be expected, of the same 
order of magnitude as those found by Rommel and Phillips. The reductions given here were 
originally made for the purjjose of comparison with other data then being studied, but it seems 
worth while to put these constants on record. To do so is the purpose of this paper. 
The data for a single year for each breed were picked out and submitted to analysis. In each 
case the year selected was that which contained the largest number of litters. For the Poland 
China the data given by the American Poland China Record for 1902 were used. The luimber of 
litters recorded in this year was 6,627. In the case of the Duroc Jersey breed the records given 
by the National Duroc Jersey Record for the same year, 1902, were used. The number of litters 
recorded in this case was 6,109. 
* Papers from the Biological Laboratory of the Maine Agricultural Experiment Station, No. 3. 
t Circular, No. 95. Bureau of Animal Industry, U. S. Department of Agriculture, Washington, D.C. 
1906. 
J Proceedings of American Philosophical Society, Vol. xlv. pp. 245 — 254, 1907, 
Biometrika ti 55 
